Research on wavelet denoising method of vehicle based MEMS accelerator signal

The analysis of vehicle based MEMS accelerometer output signal is often difficult just because of the characteristic of MEMS accelerometer, electromagnetic disturbing and the sensor is often disturbed by the vehicle vibration . Aimed at this circumstance, the paper hypothesized an acceleration signal based on the analysis of the vehicle based MEMS accelerometer output signal and added a real noise as the signal to be denoised. Selected a best wavelet in such signal denoising processing with the SNR ( Signal Noise Ratio ), error average and variance as the selection criterion. The experimentation based on the real output signal of a accelerometer proved that the wavelet denoise method is suitable for the denoising processing of such badly interfered signal , the performance is better than any other methods and the wavelet denoise method is also feasible in other circumstances.